CLINICAL STUDY
Unilateral Craniofacial Microsomia: Unrecognized Cause of Pediatric Obstructive Sleep Apnea Caroline Szpalski, MD, PhD, Meredith Vandegrift, MD, Parit A. Patel, MD, Geoffrey Appelboom, MD, PhD,y Mark Fisher, MD,z Jeffrey Marcus, MD,z Joseph G. McCarthy, MD, Pradip R. Shetye, DDS, MDS, and Stephen M. Warren, MD Abstract: Bilateral craniofacial microsomia causes obstructive sleep apnea (OSA). We hypothesize that unilateral craniofacial microsomia (UCFM) is an underappreciated cause of OSA. The records of all pediatric UCFM patients from 1990 to 2010 were reviewed; only complete records were included in the study. UCFM patients with OSA (apnea hypopnea index >1/hr) were compared to UCFM patients without OSA. Univariate and multivariate Fisher and x2 tests were performed. Of the 62 UCFM patients, 7 (11.3%) had OSA. All OSA patients had Pruzansky IIB or III mandibles. OSA patients presented with snoring (71.4%), failure to thrive (FTT) (57.1%), and chronic respiratory infections (42.8%). Snoring (P < 0.001), Goldenhar syndrome (P ¼ 0.001), and FTT (P ¼ 0.004) were significantly associated with OSA, but race, obesity, clefts, respiratory anomalies, adenotonsillar hypertrophy, and laterality were not. The prevalence of OSA in UCFM patients is up to 10 times greater than in the general population. Snoring, Goldenhar syndrome, and FTT are significantly associated with the presence of OSA. Key Words: Unilateral craniofacial microsomia, hemifacial microsomia, pediatric obstructive sleep apnea, sleep disordered breathing, craniofacial syndromes, apnea (J Craniofac Surg 2015;26: 1277–1282)
U
nilateral craniofacial microsomia (UCFM), also known as hemifacial microsomia, first and second branchial arch syndrome, or oculoauricular vertebral syndrome, is the second most common congenital malformation of the head and neck (one in 5000 births) next to cleft lip and palate.1 Although the etiopathogenesis of UCFM is unknown, the consequences of first pharyngeal arch hypoplasia are well known and include malformations of the From the Department of Plastic Surgery, New York University Langone Medical Center, New York, NY; yDepartment of Neurological Surgery/ The Neurological Institute, Columbia University College of Physicians and Surgeons, New York, NY; and zDepartment of Surgery, Division of Plastic, Reconstructive, Maxillofacial and Oral Surgery, Duke University Medical Center, Durham, NC. Received June 25, 2014. Accepted for publication January 5, 2015. Address correspondence and reprint requests to Stephen M. Warren, MD, FACS, 875 Park Avenue, New York, NY 10075; E-mail: [email protected] Presented in part at the European Association of Plastic Surgeons Meeting (May 2012), the Northeastern Society of Plastic Surgeons Meeting (September 2011), the International Society of Craniofacial Surgery Meeting (August 2013), and the American Society of Plastic Surgeons Meeting (September 2011 and September 2013). The authors report no conflicts of interest. Copyright # 2015 by Mutaz B. Habal, MD ISSN: 1049-2275 DOI: 10.1097/SCS.0000000000001551
The Journal of Craniofacial Surgery
maxilla, mandible, temporomandibular joint (TMJ), muscles of mastication, ear, and, occasionally, the facial nerve. Phenotypically, UCFM patients are heterogeneous and the dysmorphology ranges from mild to severe. Unlike UCFM, bilateral hypoplasia of first branchial arch structures is a well-known cause of sleep-disordered breathing (SDB).2,3 Clinicians suspect that bilateral maxillary hypoplasia in conjunction with adenotonsillar hypertrophy or bilateral mandibular hypoplasia with glossoptosis are contributing factors to OSA in this patient population. SDB describes a group of disorders characterized by abnormalities of respiratory pattern or the quantity of ventilation during sleep.4–7 Obstructive sleep apnea (OSA), the most common SDB subgroup, is characterized by the repetitive collapse or partial collapse of the pharyngeal airway during sleep and the need to arouse to resume ventilation.8,9 In the general population, pediatric OSA is thought to be due to adenotonsillar hypertrophy, decrease in neuromuscular tone, or even obesity.4 Pediatric OSA patients present with a variety of nonspecific symptoms (Table 1).9– 12 Although the symptoms may be ambiguous, the consequences of unrecognized OSA clearly cause cardiovascular and neurobehavioral morbidities (Table 1).4 Based on previously published studies, we commonly look for OSA in patients with craniofacial dysostosis syndromes or bilateral craniofacial microsomia; however, the incidence of OSA in patients with UCFM is unknown.2,3,13– 15 Based on anecdotal observations that some UCFM patients had OSA, in the present study, we aimed to systematically investigate the incidence of OSA in UCFM patients in our patient population and to determine significant associations.
PATIENTS AND METHODS Data Collection After obtaining IRB approval (IRB#08-676), a retrospective review of all pediatric patients diagnosed with UCFM in our institution between 1990 and 2010 was conducted. Only patients with complete data were included (2 patients with UCFM and OSA were excluded from the study because their sleep studies were performed at another facility; 4 patients with UCFM alone were excluded from the study due to incomplete data). The UCFM patients were divided into 2 cohorts: those suffering OSA (apnea hypopnea index >1/hr) and those not suffering OSA. Patients were analyzed based on sex, race, obesity, cleft lip or palate, upper respiratory complications, failure to thrive, presence of adenotonsillar hypertrophy and side (laterality) of UCFM, presence of snoring, and presence of Goldenhar syndrome. Analyses were conducted on preoperative data.
UCFM Diagnosis The diagnosis of UCFM was based on clinical signs and symptoms as well as radiographic signs of the craniofacial skeleton.
Volume 26, Number 4, June 2015
1277
Copyright © 2015 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
Szpalski et al
The Journal of Craniofacial Surgery
Volume 26, Number 4, June 2015
TABLE 1. Pediatric Obstructive Sleep Apnea Signs, Symptoms, and Morbidities Signs and Symptoms Sleep disturbances Snoringþ Loud breathingþ Agitated sleep with frequent postural changes Excessive sweating during sleep Abnormal sleep positions Night terror Sleepwalking/sleep talking Enuresis12 Mouth breathing/paradoxical breathing Pauses in breathing Chest wall retractions Excessive daytime sleepiness Excessive bruxism Failure to thriveþ, poor growth, or poor appetite Neurologic/activity symptoms Hyperactivity or inattention Excessive daytime sleepiness Morning headache
Chronic respiratory infectionsþ þ
Morbidities Cardiovascular morbidities Pulmonary hypertension22–26 Arterial hypertension22–26 Cor pulmonale with heart failure and pulmonary hypertension Blood elevation27 Autonomic dysfunction28, 29
Reduced somatic growth12,24–26 Behavioral morbidities Learning and cognitive deficits with poor academic achievement30,31 Lower general intelligence Decreased language and verbal skills Emotional instability Diminished visual and auditory attention/ADHD/ Hyperactivity disorder32 Nocturnal enuresis Various behavioral problems (social withdrawal or aggression)33–36 Systemic inflammation
Presenting symptoms in our cohort of patients with UCFM and OSA.
Predominant symptoms in preadolescent children with OSA. ADHD, attention deficit-hyperactivity disorder.
Frontal and basilar view photographs were reviewed to assess for facial asymmetry and characteristic facial dysmorphology related to UCFM. For the purpose of this study, the radiographic mandibular morphology was characterized by the Mulliken-modified Pruzansky classification scheme. The Pruzansky classification scheme has 3 groups. Type I represents a mandible that has normal dimensions of the ramus and condyle; however, it is smaller in size. Mulliken modified Pruzansky’s Type II into 2 subgroups. In the Type IIA subgroup, the ramus and condyle are decreased in both overall dimension and size. Also, the relationship of the temporomandibular joint (TMJ) and the glenoid fossa is in normal anatomical alignment, and the function of the TMJ is near normal. In contrast, in the Type IIB subgroup, the ramus and condyle are decreased in overall dimension and size; however, the TMJ is malformed and medially displaced relative to normal anatomic alignment. Lastly, in the Type III group, the ramus and condyle are absent and there is no evidence of a TMJ. Goldenhar syndrome or oculo-auriculo-vertebral dysplasia (OAV) is a rare condition characterized by the triad (usually unilateral) of craniofacial microsomia, ocular dermoid cysts, and spinal anomalies.16 It is considered to be a variant of UCFM and can be present in as many as 10% of patients with UCFM.16 Diagnosis was made based on the definition of the syndrome as described by Goldenhar in 1952.17
Obstructive Sleep Apnea Diagnosis Multiple studies have looked at the effectiveness of history and physical examination in differentiating OSA from snoring and found that neither is effective in reliably distinguishing the two.4 The gold standard for diagnosis is an in-laboratory nighttime polysomnogram (PSG) as recommended by the American Academy
1278
of Sleep Medicine (AASM, http://www.aasmnet.org/practiceguidelines.aspx). Apnea was defined as a complete cessation of airflow for more then 2 breaths and hypopnea as a 50% reduction in respiratory airflow accompanied by a decrease of 3% of SaO2. An OSA diagnosis was made with a positive PSG (apnea hypopnea index [AHI] >1/hr; a normal AHI for healthy children is considered to be 0.1 0.1/hr,4 an AHI of 1-5 is considered mildly increased, 5-10 is moderately increased, 10-20 is severely increased, and greater than 20 is severely abnormal), as recommended by the AASM. Only patients with a positive PSG done in our facility were included in the study. The average arterial oxygen saturation, pulse rate, and number of desaturations were recorded for each patient. It is important to note that children have a faster respiratory rate than adults; therefore, desaturation and hypopnea are scored differently than in adults (they have shorter apnea periods). Oximetry was measured because children tend to move frequently during sleep. Therefore, the monitoring of the pulse waveform in addition to the saturation value is helpful in distinguishing motion artifact from true desaturation. Desaturations were important to measure because in some children (especially the very young ones), a pattern of persistent partial upper airway obstruction associated with desaturation, rather than obstructive apneas or hypopneas, can be found and should raise awareness.18
Statistical Analysis Univariate analyses of pertinent clinical variables with respect to UCFM and OSA were conducted using x2 test and Fisher exact tests where appropriate. When appropriate, data were analyzed by Student t test using SPSS version 18 (Chicago, IL). Univariate logistic regression was used to identify factors related to the presence of OSA in our UCFM population. The following variables #
2015 Mutaz B. Habal, MD
Copyright © 2015 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery
Volume 26, Number 4, June 2015
UCFM and OSA
TABLE 2. Population Description
Female Male Age range Age at diagnosis range AHI range AHI average Pruzansky grade I Pruzansky grade IIa Pruzansky grade IIb Pruzansky grade III
UCFM With OSA (7)
UCFM Without OSA (55)
2 5 6 months to 156 months 6 months to 84 months 2.6–20 9.9 6.8 — — 2 5
20 35
— — 50 4 — 1
At the time of the study.
were assessed: race, obesity, cleft lip or palate, upper respiratory complications, failure to thrive, presence of adenotonsillar hypertrophy and side (laterality) of UCFM, presence of snoring, and Goldenhar syndrome. Using the same variables, multiple logistic regression analysis was used to identify significant independent relationships to OSA. A P value of 0.05 was considered significant. All analyses were performed with SPSS version 18 (Chicago, IL).
RESULTS Population Description OSA Occurs in Patients With UCFM Sixty-two patients were identified as having UCFM (clinical and radiographic diagnosis). Forty males and twenty-two females were identified (Table 2). Of the 62 patients, seven were diagnosed with OSA (AHI range ¼ 2.6-20, average AHI ¼ 9.9 6.8). Of the 7 patients diagnosed with OSA, 3 (42%) had severe OSA, 2 (28%) had moderate OSA, and 2 (28%) had mild OSA (Table 3). The average pulse rate for the severe, moderate, and mild OSA groups was 89 beats per minute (bpm), 115 bpm, and 83 bpm, respectively. The average oxygen saturation for the severe, moderate, and mild OSA groups was 95%, 97%, and 97%, respectively. The average number of desaturations for the severe, moderate, and mild OSA groups was 11.6, 8.5, and 6.5, respectively. According to some studies, the male gender predominates in the OSA population (especially in older pediatric populations).4 In our study, males outnumbered females in subject numbers (Fig. 1); however, there was no statistical difference in terms of OSA prevalence between female and male patients (Fig. 1).
FIGURE 1. Sex distribution. Although male sex is a known risk factor for obstructive sleep apnea and there were more males than females in our cohort, there was no significant difference the incidence of obstructive sleep apnea in males and females. OSA, obstructive sleep apnea; UCFM, unilateral craniofacial microsomia; NS, not significant.
Pruzansky grade I or IIA (P ¼ 0.001) (Table 2). Of note, 1 Pruzansky grade III patient out of the 6 did not have OSA.
Ethnicity is Not Related to the Presence of OSA in UCFM Patients Because race can be a risk factor for OSA in the otherwise healthy population (African American children are at an increased risk because of both their craniofacial structure and socioeconomic status),2,5– 19 we also looked at the ethnicity of our population (Fig. 2). In our cohort, 69.4% of the population was Caucasian, 22.6% were Hispanic, 4.8% were African American, and 3.2% of our population was Asian. Race was not a significant factor when the affected group (UCFM þ OSA) and the non-affected group were compared (P ¼ 0.2).
Snoring is the Most Common Presenting Symptom of OSA and is Associated With OSA in UCFM Patients In our cohort, patients with OSA presented with various symptoms including snoring (71.4%), failure to thrive (FTT) (57.1%),
UCFM Patients With OSA Have More Severe Forms of Craniofacial Microsomia All patients with OSA presented with Pruzansky grades IIB or higher whereas 98% of UCFM patients who did not have OSA had a TABLE 3. Polysomnography Results
Number Average pulse Average O2 saturation Average number of desaturations/hr
#
2015 Mutaz B. Habal, MD
Mild OSA
Moderate OSA
Severe OSA
2 (28%) 83 bpm 97% 6.5
2 (28%) 115 bpm 97% 8.5
3 (42%) 89 bpm 95% 11.6
FIGURE 2. Ethnicity distribution. Race is a known risk factor for obstructive sleep apnea (eg, African American children are at a higher risk for obstructive sleep apnea). In our study, 69.4% of the population was Caucasian, 22.6% were Hispanic, 4.8% were African American, and 3.2% of our population was Asian. Race was not a significant factor when comparing the obstructive sleep apnea group to non-affected group (P ¼ 0.2). OSA, obstructive sleep apnea; UCFM, unilateral craniofacial microsomia; NS, not significant.
1279
Copyright © 2015 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery
Szpalski et al
Volume 26, Number 4, June 2015
TABLE 5. Multivariate Analysis
Factor A. Design 1 Cardiac issues Small for age Goldenhar syndrome Snoring B. Design 2 Cardiac issues Small for age Goldenhar syndrome FIGURE 3. Presenting symptoms in our cohort. In our cohort, patients with obstructive sleep apnea presented with various symptoms. Snoring was by far the most common presenting symptom (71.4%), followed by failure to thrive (57.1%), chronic respiratory infections (42.8%), adenotonsillar hypertrophy (28.6%), and finally loud breathing (28.5%).
chronic respiratory infections (42.8%), adenotonsillar hypertrophy (28.6%), or loud breathing (28.5%) (Table 1, Fig. 3).
Snoring, Diagnosis of Goldenhar Syndrome, and Failure to Thrive are Significantly Related to the Presence of OSA in UCFM Patients To identify factors related to OSA in the UCFM population, a univariate test was created. Snoring (P < 0.001), presence of Goldenhar syndrome (P ¼ 0.001), and FTT (P ¼ 0.004) were significantly correlated with the presence of OSA in patients with UCFM (Table 4). Sex, side of the microsomia, ethnicity, and presence of cardiac issues were not related to the presence of OSA. More importantly, obesity, an important risk factor in the otherwise healthy population, was not correlated with the presence of OSA and was therefore not identified as a risk factor in our cohort (P ¼ 0.113).
Snoring is Independently Associated With OSA To define independent association for OSA in that same population, multivariate analysis models were created (Table 5). In our first model, we calculated the odds ratio of cardiac issues, FTT, presence of Goldenhar syndrome, and presence of snoring in the 2
TABLE 4. Univariate Analysis
Factors Female Ethnicity Caucasian African-American Asian Hispanic Obesity Cardiac problems Snoring Left-sided microstomia Small for age Cleft lip/palate Goldenhar syndrome
1280
Cohort (n ¼ 62)
OSA (n ¼ 7)
No OSA (n ¼ 55)
20 (32.3)
2 (28.6)
18 (32.7)
43 3 2 14 1 10 9 29 8 11 4
5 (71,4) 1 (14.3) 0 1 (14.3) 1 (14.3) 3 (42.9) 5 (71.4) 2 (28.6) 4 (57.1) 5 (71.4) 1 (14.3)
38 2 2 13
(69.4) (4.8) (3.2) (22.6) (1.6) (16.1) (14.5) (46.8) (12.9) (17.7) (6.5)
7 4 27 4 6 3
(69.1) (3.6) (28.6) (23.6) 0 (12.7) (7.3) (49.1) (7.3) (10.9) (5.5)
P Value 1 0.2
0.113 0.0076
Unilateral Craniofacial Microsomia: Unrecognized Cause of Pediatric Obstructive Sleep Apnea Caroline Szpalski, MD, PhD, Meredith Vandegrift, MD, Parit A. Patel, MD, Geoffrey Appelboom, MD, PhD,y Mark Fisher, MD,z Jeffrey Marcus, MD,z Joseph G. McCarthy, MD, Pradip R. Shetye, DDS, MDS, and Stephen M. Warren, MD Abstract: Bilateral craniofacial microsomia causes obstructive sleep apnea (OSA). We hypothesize that unilateral craniofacial microsomia (UCFM) is an underappreciated cause of OSA. The records of all pediatric UCFM patients from 1990 to 2010 were reviewed; only complete records were included in the study. UCFM patients with OSA (apnea hypopnea index >1/hr) were compared to UCFM patients without OSA. Univariate and multivariate Fisher and x2 tests were performed. Of the 62 UCFM patients, 7 (11.3%) had OSA. All OSA patients had Pruzansky IIB or III mandibles. OSA patients presented with snoring (71.4%), failure to thrive (FTT) (57.1%), and chronic respiratory infections (42.8%). Snoring (P < 0.001), Goldenhar syndrome (P ¼ 0.001), and FTT (P ¼ 0.004) were significantly associated with OSA, but race, obesity, clefts, respiratory anomalies, adenotonsillar hypertrophy, and laterality were not. The prevalence of OSA in UCFM patients is up to 10 times greater than in the general population. Snoring, Goldenhar syndrome, and FTT are significantly associated with the presence of OSA. Key Words: Unilateral craniofacial microsomia, hemifacial microsomia, pediatric obstructive sleep apnea, sleep disordered breathing, craniofacial syndromes, apnea (J Craniofac Surg 2015;26: 1277–1282)
U
nilateral craniofacial microsomia (UCFM), also known as hemifacial microsomia, first and second branchial arch syndrome, or oculoauricular vertebral syndrome, is the second most common congenital malformation of the head and neck (one in 5000 births) next to cleft lip and palate.1 Although the etiopathogenesis of UCFM is unknown, the consequences of first pharyngeal arch hypoplasia are well known and include malformations of the From the Department of Plastic Surgery, New York University Langone Medical Center, New York, NY; yDepartment of Neurological Surgery/ The Neurological Institute, Columbia University College of Physicians and Surgeons, New York, NY; and zDepartment of Surgery, Division of Plastic, Reconstructive, Maxillofacial and Oral Surgery, Duke University Medical Center, Durham, NC. Received June 25, 2014. Accepted for publication January 5, 2015. Address correspondence and reprint requests to Stephen M. Warren, MD, FACS, 875 Park Avenue, New York, NY 10075; E-mail: [email protected] Presented in part at the European Association of Plastic Surgeons Meeting (May 2012), the Northeastern Society of Plastic Surgeons Meeting (September 2011), the International Society of Craniofacial Surgery Meeting (August 2013), and the American Society of Plastic Surgeons Meeting (September 2011 and September 2013). The authors report no conflicts of interest. Copyright # 2015 by Mutaz B. Habal, MD ISSN: 1049-2275 DOI: 10.1097/SCS.0000000000001551
The Journal of Craniofacial Surgery
maxilla, mandible, temporomandibular joint (TMJ), muscles of mastication, ear, and, occasionally, the facial nerve. Phenotypically, UCFM patients are heterogeneous and the dysmorphology ranges from mild to severe. Unlike UCFM, bilateral hypoplasia of first branchial arch structures is a well-known cause of sleep-disordered breathing (SDB).2,3 Clinicians suspect that bilateral maxillary hypoplasia in conjunction with adenotonsillar hypertrophy or bilateral mandibular hypoplasia with glossoptosis are contributing factors to OSA in this patient population. SDB describes a group of disorders characterized by abnormalities of respiratory pattern or the quantity of ventilation during sleep.4–7 Obstructive sleep apnea (OSA), the most common SDB subgroup, is characterized by the repetitive collapse or partial collapse of the pharyngeal airway during sleep and the need to arouse to resume ventilation.8,9 In the general population, pediatric OSA is thought to be due to adenotonsillar hypertrophy, decrease in neuromuscular tone, or even obesity.4 Pediatric OSA patients present with a variety of nonspecific symptoms (Table 1).9– 12 Although the symptoms may be ambiguous, the consequences of unrecognized OSA clearly cause cardiovascular and neurobehavioral morbidities (Table 1).4 Based on previously published studies, we commonly look for OSA in patients with craniofacial dysostosis syndromes or bilateral craniofacial microsomia; however, the incidence of OSA in patients with UCFM is unknown.2,3,13– 15 Based on anecdotal observations that some UCFM patients had OSA, in the present study, we aimed to systematically investigate the incidence of OSA in UCFM patients in our patient population and to determine significant associations.
PATIENTS AND METHODS Data Collection After obtaining IRB approval (IRB#08-676), a retrospective review of all pediatric patients diagnosed with UCFM in our institution between 1990 and 2010 was conducted. Only patients with complete data were included (2 patients with UCFM and OSA were excluded from the study because their sleep studies were performed at another facility; 4 patients with UCFM alone were excluded from the study due to incomplete data). The UCFM patients were divided into 2 cohorts: those suffering OSA (apnea hypopnea index >1/hr) and those not suffering OSA. Patients were analyzed based on sex, race, obesity, cleft lip or palate, upper respiratory complications, failure to thrive, presence of adenotonsillar hypertrophy and side (laterality) of UCFM, presence of snoring, and presence of Goldenhar syndrome. Analyses were conducted on preoperative data.
UCFM Diagnosis The diagnosis of UCFM was based on clinical signs and symptoms as well as radiographic signs of the craniofacial skeleton.
Volume 26, Number 4, June 2015
1277
Copyright © 2015 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
Szpalski et al
The Journal of Craniofacial Surgery
Volume 26, Number 4, June 2015
TABLE 1. Pediatric Obstructive Sleep Apnea Signs, Symptoms, and Morbidities Signs and Symptoms Sleep disturbances Snoringþ Loud breathingþ Agitated sleep with frequent postural changes Excessive sweating during sleep Abnormal sleep positions Night terror Sleepwalking/sleep talking Enuresis12 Mouth breathing/paradoxical breathing Pauses in breathing Chest wall retractions Excessive daytime sleepiness Excessive bruxism Failure to thriveþ, poor growth, or poor appetite Neurologic/activity symptoms Hyperactivity or inattention Excessive daytime sleepiness Morning headache
Chronic respiratory infectionsþ þ
Morbidities Cardiovascular morbidities Pulmonary hypertension22–26 Arterial hypertension22–26 Cor pulmonale with heart failure and pulmonary hypertension Blood elevation27 Autonomic dysfunction28, 29
Reduced somatic growth12,24–26 Behavioral morbidities Learning and cognitive deficits with poor academic achievement30,31 Lower general intelligence Decreased language and verbal skills Emotional instability Diminished visual and auditory attention/ADHD/ Hyperactivity disorder32 Nocturnal enuresis Various behavioral problems (social withdrawal or aggression)33–36 Systemic inflammation
Presenting symptoms in our cohort of patients with UCFM and OSA.
Predominant symptoms in preadolescent children with OSA. ADHD, attention deficit-hyperactivity disorder.
Frontal and basilar view photographs were reviewed to assess for facial asymmetry and characteristic facial dysmorphology related to UCFM. For the purpose of this study, the radiographic mandibular morphology was characterized by the Mulliken-modified Pruzansky classification scheme. The Pruzansky classification scheme has 3 groups. Type I represents a mandible that has normal dimensions of the ramus and condyle; however, it is smaller in size. Mulliken modified Pruzansky’s Type II into 2 subgroups. In the Type IIA subgroup, the ramus and condyle are decreased in both overall dimension and size. Also, the relationship of the temporomandibular joint (TMJ) and the glenoid fossa is in normal anatomical alignment, and the function of the TMJ is near normal. In contrast, in the Type IIB subgroup, the ramus and condyle are decreased in overall dimension and size; however, the TMJ is malformed and medially displaced relative to normal anatomic alignment. Lastly, in the Type III group, the ramus and condyle are absent and there is no evidence of a TMJ. Goldenhar syndrome or oculo-auriculo-vertebral dysplasia (OAV) is a rare condition characterized by the triad (usually unilateral) of craniofacial microsomia, ocular dermoid cysts, and spinal anomalies.16 It is considered to be a variant of UCFM and can be present in as many as 10% of patients with UCFM.16 Diagnosis was made based on the definition of the syndrome as described by Goldenhar in 1952.17
Obstructive Sleep Apnea Diagnosis Multiple studies have looked at the effectiveness of history and physical examination in differentiating OSA from snoring and found that neither is effective in reliably distinguishing the two.4 The gold standard for diagnosis is an in-laboratory nighttime polysomnogram (PSG) as recommended by the American Academy
1278
of Sleep Medicine (AASM, http://www.aasmnet.org/practiceguidelines.aspx). Apnea was defined as a complete cessation of airflow for more then 2 breaths and hypopnea as a 50% reduction in respiratory airflow accompanied by a decrease of 3% of SaO2. An OSA diagnosis was made with a positive PSG (apnea hypopnea index [AHI] >1/hr; a normal AHI for healthy children is considered to be 0.1 0.1/hr,4 an AHI of 1-5 is considered mildly increased, 5-10 is moderately increased, 10-20 is severely increased, and greater than 20 is severely abnormal), as recommended by the AASM. Only patients with a positive PSG done in our facility were included in the study. The average arterial oxygen saturation, pulse rate, and number of desaturations were recorded for each patient. It is important to note that children have a faster respiratory rate than adults; therefore, desaturation and hypopnea are scored differently than in adults (they have shorter apnea periods). Oximetry was measured because children tend to move frequently during sleep. Therefore, the monitoring of the pulse waveform in addition to the saturation value is helpful in distinguishing motion artifact from true desaturation. Desaturations were important to measure because in some children (especially the very young ones), a pattern of persistent partial upper airway obstruction associated with desaturation, rather than obstructive apneas or hypopneas, can be found and should raise awareness.18
Statistical Analysis Univariate analyses of pertinent clinical variables with respect to UCFM and OSA were conducted using x2 test and Fisher exact tests where appropriate. When appropriate, data were analyzed by Student t test using SPSS version 18 (Chicago, IL). Univariate logistic regression was used to identify factors related to the presence of OSA in our UCFM population. The following variables #
2015 Mutaz B. Habal, MD
Copyright © 2015 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery
Volume 26, Number 4, June 2015
UCFM and OSA
TABLE 2. Population Description
Female Male Age range Age at diagnosis range AHI range AHI average Pruzansky grade I Pruzansky grade IIa Pruzansky grade IIb Pruzansky grade III
UCFM With OSA (7)
UCFM Without OSA (55)
2 5 6 months to 156 months 6 months to 84 months 2.6–20 9.9 6.8 — — 2 5
20 35
— — 50 4 — 1
At the time of the study.
were assessed: race, obesity, cleft lip or palate, upper respiratory complications, failure to thrive, presence of adenotonsillar hypertrophy and side (laterality) of UCFM, presence of snoring, and Goldenhar syndrome. Using the same variables, multiple logistic regression analysis was used to identify significant independent relationships to OSA. A P value of 0.05 was considered significant. All analyses were performed with SPSS version 18 (Chicago, IL).
RESULTS Population Description OSA Occurs in Patients With UCFM Sixty-two patients were identified as having UCFM (clinical and radiographic diagnosis). Forty males and twenty-two females were identified (Table 2). Of the 62 patients, seven were diagnosed with OSA (AHI range ¼ 2.6-20, average AHI ¼ 9.9 6.8). Of the 7 patients diagnosed with OSA, 3 (42%) had severe OSA, 2 (28%) had moderate OSA, and 2 (28%) had mild OSA (Table 3). The average pulse rate for the severe, moderate, and mild OSA groups was 89 beats per minute (bpm), 115 bpm, and 83 bpm, respectively. The average oxygen saturation for the severe, moderate, and mild OSA groups was 95%, 97%, and 97%, respectively. The average number of desaturations for the severe, moderate, and mild OSA groups was 11.6, 8.5, and 6.5, respectively. According to some studies, the male gender predominates in the OSA population (especially in older pediatric populations).4 In our study, males outnumbered females in subject numbers (Fig. 1); however, there was no statistical difference in terms of OSA prevalence between female and male patients (Fig. 1).
FIGURE 1. Sex distribution. Although male sex is a known risk factor for obstructive sleep apnea and there were more males than females in our cohort, there was no significant difference the incidence of obstructive sleep apnea in males and females. OSA, obstructive sleep apnea; UCFM, unilateral craniofacial microsomia; NS, not significant.
Pruzansky grade I or IIA (P ¼ 0.001) (Table 2). Of note, 1 Pruzansky grade III patient out of the 6 did not have OSA.
Ethnicity is Not Related to the Presence of OSA in UCFM Patients Because race can be a risk factor for OSA in the otherwise healthy population (African American children are at an increased risk because of both their craniofacial structure and socioeconomic status),2,5– 19 we also looked at the ethnicity of our population (Fig. 2). In our cohort, 69.4% of the population was Caucasian, 22.6% were Hispanic, 4.8% were African American, and 3.2% of our population was Asian. Race was not a significant factor when the affected group (UCFM þ OSA) and the non-affected group were compared (P ¼ 0.2).
Snoring is the Most Common Presenting Symptom of OSA and is Associated With OSA in UCFM Patients In our cohort, patients with OSA presented with various symptoms including snoring (71.4%), failure to thrive (FTT) (57.1%),
UCFM Patients With OSA Have More Severe Forms of Craniofacial Microsomia All patients with OSA presented with Pruzansky grades IIB or higher whereas 98% of UCFM patients who did not have OSA had a TABLE 3. Polysomnography Results
Number Average pulse Average O2 saturation Average number of desaturations/hr
#
2015 Mutaz B. Habal, MD
Mild OSA
Moderate OSA
Severe OSA
2 (28%) 83 bpm 97% 6.5
2 (28%) 115 bpm 97% 8.5
3 (42%) 89 bpm 95% 11.6
FIGURE 2. Ethnicity distribution. Race is a known risk factor for obstructive sleep apnea (eg, African American children are at a higher risk for obstructive sleep apnea). In our study, 69.4% of the population was Caucasian, 22.6% were Hispanic, 4.8% were African American, and 3.2% of our population was Asian. Race was not a significant factor when comparing the obstructive sleep apnea group to non-affected group (P ¼ 0.2). OSA, obstructive sleep apnea; UCFM, unilateral craniofacial microsomia; NS, not significant.
1279
Copyright © 2015 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery
Szpalski et al
Volume 26, Number 4, June 2015
TABLE 5. Multivariate Analysis
Factor A. Design 1 Cardiac issues Small for age Goldenhar syndrome Snoring B. Design 2 Cardiac issues Small for age Goldenhar syndrome FIGURE 3. Presenting symptoms in our cohort. In our cohort, patients with obstructive sleep apnea presented with various symptoms. Snoring was by far the most common presenting symptom (71.4%), followed by failure to thrive (57.1%), chronic respiratory infections (42.8%), adenotonsillar hypertrophy (28.6%), and finally loud breathing (28.5%).
chronic respiratory infections (42.8%), adenotonsillar hypertrophy (28.6%), or loud breathing (28.5%) (Table 1, Fig. 3).
Snoring, Diagnosis of Goldenhar Syndrome, and Failure to Thrive are Significantly Related to the Presence of OSA in UCFM Patients To identify factors related to OSA in the UCFM population, a univariate test was created. Snoring (P < 0.001), presence of Goldenhar syndrome (P ¼ 0.001), and FTT (P ¼ 0.004) were significantly correlated with the presence of OSA in patients with UCFM (Table 4). Sex, side of the microsomia, ethnicity, and presence of cardiac issues were not related to the presence of OSA. More importantly, obesity, an important risk factor in the otherwise healthy population, was not correlated with the presence of OSA and was therefore not identified as a risk factor in our cohort (P ¼ 0.113).
Snoring is Independently Associated With OSA To define independent association for OSA in that same population, multivariate analysis models were created (Table 5). In our first model, we calculated the odds ratio of cardiac issues, FTT, presence of Goldenhar syndrome, and presence of snoring in the 2
TABLE 4. Univariate Analysis
Factors Female Ethnicity Caucasian African-American Asian Hispanic Obesity Cardiac problems Snoring Left-sided microstomia Small for age Cleft lip/palate Goldenhar syndrome
1280
Cohort (n ¼ 62)
OSA (n ¼ 7)
No OSA (n ¼ 55)
20 (32.3)
2 (28.6)
18 (32.7)
43 3 2 14 1 10 9 29 8 11 4
5 (71,4) 1 (14.3) 0 1 (14.3) 1 (14.3) 3 (42.9) 5 (71.4) 2 (28.6) 4 (57.1) 5 (71.4) 1 (14.3)
38 2 2 13
(69.4) (4.8) (3.2) (22.6) (1.6) (16.1) (14.5) (46.8) (12.9) (17.7) (6.5)
7 4 27 4 6 3
(69.1) (3.6) (28.6) (23.6) 0 (12.7) (7.3) (49.1) (7.3) (10.9) (5.5)
P Value 1 0.2
0.113 0.0076
